Tim Dettmers Calls Out Overstated AI Security Risks: GPT-2 Could Hack HF Backend

Tim_Dettmers · x · 2026-07-31

Tim Dettmers, well-known AI researcher, questions the reality of current AI security risks. He mentioned that security researchers found the "Mythos" risks somewhat fake, noting AI was just used to fix security issues they previously lacked time for. He added that even his grandma could have hacked Hugging Face's backend using GPT-2.

Original post →

More from Fun

Fun channel →